← 返回 Skills 市场
sidvorak

Wine Archive

作者 Simon I Dvorak · GitHub ↗ · v0.1.8 · MIT-0
cross-platform ✓ 安全检测通过
141
总下载
1
收藏
0
当前安装
9
版本数
在 OpenClaw 中安装
/install wine-archive
功能描述
Store, recall, and manage personal wine tastings and labels using natural language queries with durable image storage in a local SQLite database.
安全使用建议
This skill appears to be a straightforward local wine archive. Before installing, consider: 1) Inspect setup.sh—it runs npm install and may build native modules (better-sqlite3/prebuild-install downloads are normal for native packages). Run it in an environment you control. 2) Node >= 22 is required; the registry metadata did not declare this—make sure your runtime meets it. 3) Telegram features require the OpenClaw CLI and a configured Telegram bot; those are optional but not declared as required binaries. 4) The skill writes a SQLite DB and copies images to data/wine/labels (default path) and appends to data/wine/audit.log — back up any existing data and review WINE_DB_PATH if you want a custom location. 5) LLM features are optional and require you to add ANTHROPIC_API_KEY or OPENAI_API_KEY to a .env; do not provide keys unless you want cloud LLM classification. 6) Export/Import can embed images as base64 in a single JSON file (sensitive data) — be careful sharing exports. If you want extra caution, run the setup and npm install inside a disposable container or VM and review the contents of setup.sh and any scripts before executing them.
功能分析
Type: OpenClaw Skill Name: wine-archive Version: 0.1.8 The wine-archive skill is a legitimate personal database for tracking wine tastings and labels. It follows security best practices, such as using parameterized queries in lib/wine-store.js to prevent SQL injection, implementing secret redaction for LLM interaction logs in shared/interaction-store.js, and using shell quoting for generated CLI commands in scripts/wine-chat.js. The functionality, including data export/import and Telegram integration, is well-documented and aligns with the stated purpose without any evidence of malicious intent or data exfiltration.
能力标签
cryptocan-make-purchasesrequires-oauth-tokenrequires-sensitive-credentials
能力评估
Purpose & Capability
The code and instructions align with the described purpose (local SQLite DB for wines, image archival, optional LLM intent classification, Telegram bridge). Minor inconsistencies: the registry metadata lists no required binaries or env vars, but README/SKILL.md require Node.js >=22 and (optionally) ANTHROPIC_API_KEY / OPENAI_API_KEY and OpenClaw CLI for Telegram features; those runtime prerequisites are reasonable for the declared features but are not reflected in the simple 'required' metadata.
Instruction Scope
SKILL.md and the scripts focus on creating/reading a local DB, archiving label images into data/wine/labels, exporting/importing JSON (optionally embedding images), and optionally calling an LLM or the OpenClaw CLI for Telegram. The instructions do ask you to run setup.sh and npm install and will write files under the skill's data directory and the SQLite DB; they do not (in the visible code) instruct reading or exfiltrating unrelated system files or secrets.
Install Mechanism
No install spec in registry (instruction-only), but repository contains Node.js code and a setup.sh that runs npm install. Dependencies come from npm (better-sqlite3, @anthropic-ai/sdk) — these are common and resolved via package-lock.json. better-sqlite3 is a native module with an install/build phase (prebuild-install) which may fetch prebuilt binaries during npm install; this is expected but worth noting. There are no downloads from unusual personal servers or URL shorteners in the visible files.
Credentials
Environment variables mentioned are proportional to the optional features: ANTHROPIC_API_KEY / OPENAI_API_KEY only for LLM classification, WINE_DB_PATH to override DB location. The skill does not request unrelated credentials or tokens. Telegram integration requires the OpenClaw CLI (and the user’s configured Telegram bot), which is expected for that capability but not declared in the skill's registry 'required binaries' section.
Persistence & Privilege
The skill does persistent local storage (SQLite DB, label image copies, audit.log) in the skill's data directory. always:false (not force-included). It does perform migration of legacy image paths inside its own DB and writes/reads files under data/wine — this is consistent with its purpose and does not appear to modify other skills or system-wide agent settings.
如何使用
  1. 确保已安装 OpenClaw(本地或 Docker 部署)
  2. 在对话框中输入安装命令:/install wine-archive
  3. 安装完成后,直接呼叫该 Skill 的名称或使用 /wine-archive 触发
  4. 根据 Skill 的参数说明提供必要输入,即可获得结构化输出
版本历史
v0.1.8
Auto-migrate image paths from old workspace location to skill directory on database open, handle broken paths gracefully
v0.1.7
Regenerate package-lock.json to clear scanner cache
v0.1.6
Regenerate package-lock.json to clear scanner cache
v0.1.5
Remove Node crypto dependency; use pure-JS FNV-1a hash for image deduplication to avoid crypto scanner false-positive
v0.1.4
Document OPENAI_API_KEY and clarify offline-first defaults in env vars
v0.1.3
Switch to MIT license
v0.1.2
Remove child_process usage to clear security scanner flags
v0.1.1
Remove eval in Telegram helper; use execFileSync directly
v0.1.0
Initial release
元数据
Slug wine-archive
版本 0.1.8
许可证 MIT-0
累计安装 0
当前安装数 0
历史版本数 9
常见问题

Wine Archive 是什么?

Store, recall, and manage personal wine tastings and labels using natural language queries with durable image storage in a local SQLite database. 它是一个面向 Claude Code / OpenClaw 的 AI Agent Skill 插件,目前累计下载 141 次。

如何安装 Wine Archive?

在 OpenClaw 或 Claude Code 对话框中运行命令「/install wine-archive」即可一键安装,无需额外配置。

Wine Archive 是免费的吗?

是的,Wine Archive 完全免费,采用 MIT-0 许可证,可自由下载、安装和使用。

Wine Archive 支持哪些平台?

Wine Archive 跨平台运行,可在任意部署了 OpenClaw / Claude Code 的环境中使用(cross-platform)。

谁开发了 Wine Archive?

由 Simon I Dvorak(@sidvorak)开发并维护,当前版本 v0.1.8。

💬 留言讨论